Urban traffic atmospheric particulate concentration prediction method, device, equipment and medium

Abstract

The invention relates to an urban traffic atmospheric particulate concentration prediction method, device and equipment and a medium. The method comprises the following steps: acquiring a plurality of urban streetscape images, corresponding geographical location information and PM2.5 concentration of a traffic road section; calculating and determining a fluctuation ratio in PM2.5 pollution and a recovery ratio after PM2.5 pollution corresponding to the traffic road section where the urban street view image is located based on the PM2.5 concentration, and calculating and determining a PM2.5 road section elastic coefficient corresponding to the traffic road section where the urban street view image is located based on the fluctuation ratio in PM2.5 pollution and the recovery ratio after PM2.5 pollution; and constructing a sample data set based on the geographic position information, the urban environment characteristic factor, the fluctuation ratio, the recovery ratio, the PM2.5 road section elastic coefficient and the PM2.5 concentration corresponding to the PM2.5 road section elastic coefficient, and predicting the PM2.5 concentration corresponding to the traffic road section where the target urban streetscape image is located by adopting an atmospheric particulate concentration prediction model trained by the sample data set. According to the invention, the prediction precision of atmospheric particulate concentration prediction can be significantly improved.

Description

Technical Field

[0001] The present application relates to the field of air pollution detection, and in particular to a method for predicting the concentration of particulate matter in urban traffic atmosphere, a corresponding device, an electronic device and a computer-readable storage medium. Background Art

[0002] With the rapid development of society and the acceleration of urbanization, air pollution control has become a major issue facing urban development. Among the air pollutants in many cities, atmospheric particulate matter is considered to be one of the main air pollutants, which seriously threatens the sustainable development of the ecological environment and human health.

[0003] Atmospheric particle concentration monitoring is the data basis for conducting atmospheric particle concentration predictions. Current atmospheric particle concentration monitoring mainly relies on fixed monitoring stations. Due to the small number of fixed monitoring stations, limited monitoring range, and the existence of monitoring blind spots, in recent years, the integration of mobile monitoring technology into atmospheric particle concentration monitoring has become a new direction. The full coverage of mobile monitoring technology can not only provide more accurate and high-resolution atmospheric particle concentration data, but also reveal the dynamic distribution of atmospheric particles in different urban built environments.

[0004] The concentration of atmospheric particulate matter is closely related to urbanization and urban activities, including industrial production, transportation, energy consumption, etc. In addition, the distribution pattern of atmospheric particulate matter concentration is also affected by multiple factors of the urban built environment. For example, the cluster of street-facing buildings has a significant impact on the diffusion of atmospheric particulate matter; the layout and height of buildings have an impact on airflow, which will lead to an increase in the accumulation level of atmospheric particulate matter. It can be seen that the urban built environment plays a key role in the distribution pattern and concentration regulation of atmospheric particulate matter. Therefore, in-depth research on the impact of the urban built environment on the concentration of atmospheric particulate matter is of great significance for understanding and responding to urban air pollution.

[0005] Domestic and foreign scholars have carried out a lot of research on atmospheric particulate matter pollution and concentration prediction. In terms of research methods, early studies used simple linear or nonlinear regression models to estimate PM 2.5 However, these methods often fail to adequately capture the complexity of the urban built environment and the nonlinear relationships in atmospheric particulate matter concentrations.

[0006] In recent years, many studies have begun to use advanced statistical models to predict PM 2.5 Concentration, such as generalized linear model (GLM), generalized additive model (GAM) and land use regression (LUR), these traditional statistical models are based on data modeling and produce estimates with fixed structure. However, traditional fixed structure statistical models may simplify the PM 2.5Modeling of complex relationships between concentrations and explanatory variables, for example, ignoring interactions between explanatory variables or relationships between explanatory variables and PM 2.5 The nonlinear relationship between the concentrations.

[0007] Although domestic and foreign scholars have conducted multi-dimensional and multi-angle discussions and research in the field of atmospheric particulate matter pollution and concentration prediction, and proposed many relatively effective atmospheric particulate matter prediction models, these studies focus on using land use models or meteorological elements for prediction, and mainly focus on factors in the horizontal direction of the ground, such as road networks, building layouts, and land use types. There are still deficiencies and shortcomings:

[0008] First, in PM 2.5 In terms of concentration data collection methods, most studies use traditional fixed monitoring stations for data collection. However, fixed monitoring stations have problems such as a small number of stations, limited monitoring range, and monitoring blind spots, and cannot provide full-area PM 2.5 Concentration data, while mobile monitoring can obtain data from multiple points in real time, but existing mobile monitoring mainly relies on handheld mobile sensors, which require high manpower costs, or professional atmospheric cruise monitoring systems placed on taxis, which have insufficient coverage of remote areas.

[0009] Second, in predicting PM 2.5 Concentration and characterization of PM 2.5 In terms of diffusion paths, computational fluid dynamics (CFD) simulation methods are mainly used for exploration. This method simplifies the urban built environment while giving approximate rather than realistic and accurate prediction results. Although CFD models can provide vertical dimension analysis, CFD models have high hardware requirements, long simulation time, and high prediction costs.

[0010] Third, urban street view images are increasingly being used to measure the vertical dimension of urban street environments. However, urban street view images are mostly used in urban planning, and are rarely used in PM. 2.5 Tracing and predicting research, detailed description of urban built environment, in-depth exploration of the relationship between urban built environment and PM 2.5 The coupling relationship between pollution.

[0011] To sum up, in order to adapt to the problems that mobile monitoring in the existing technology mainly relies on handheld mobile sensors, which have high labor costs and insufficient coverage, and the CFD model has high hardware requirements, long simulation time, and high prediction costs, the applicant has made corresponding explorations to solve these problems. Summary of the invention

[0012] The purpose of the present application is to solve the above-mentioned problems and to provide a method for predicting the concentration of particulate matter in urban traffic atmosphere, a corresponding device, an electronic device and a computer-readable storage medium.

[0013] In order to meet the various objectives of this application, this application adopts the following technical solutions:

[0014] A method for predicting the concentration of particulate matter in urban traffic atmosphere proposed to meet one of the purposes of this application includes:

[0015] In response to the urban traffic atmospheric particulate matter concentration prediction command, multiple city street scene images taken continuously by the vehicle-mounted camera equipment and their corresponding geographical location information, PM concentration of the traffic section where they are located are obtained. 2.5 Concentration, wherein the geographic location information includes one or more of POI information, normalized vegetation index, and building density;

[0016] Performing image segmentation on the urban street view image using an image segmentation model that has been trained to a convergent state, so as to extract mask data corresponding to the contour of a target area in each target area image in the urban street view image, wherein the target area includes a green plant area, a motor vehicle area, a sky area, a building area, a fence area, a column area, and a wall area;

[0017] Calculate and determine the number of pixels corresponding to the target area contour in each target area image in the urban street view image according to the mask data, and calculate and determine the urban environment characteristic factor according to the number of pixels corresponding to the target area contour, wherein the urban environment characteristic factor includes a green view rate, a vehicle interference index, a sky visibility index, and a space enclosure index;

[0018] Based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 Fluctuation rates in pollution and PM 2.5 Recovery ratio after pollution, based on the PM 2.5 Fluctuations in pollution and the PM 2.5 The recovery ratio after pollution is calculated to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Section elasticity coefficient;

[0019] Based on the geographical location information, urban environment characteristic factors, fluctuation ratio, recovery ratio, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The concentration of the target city is used to construct a sample data set, and the atmospheric particulate matter concentration prediction model trained with the sample data set is used to predict the PM2.5 concentration of the target city corresponding to the traffic section where the street view image is located. 2.5 concentration to complete the prediction of urban traffic atmospheric particulate matter concentration.

[0020] Optionally, the step of using an image segmentation model that has been trained to a convergent state to perform image segmentation on the city street view image to extract mask data corresponding to the target area contour in each target area image in the city street view image comprises:

[0021] Performing multi-level encoding on the target area image and correspondingly generating intermediate feature information of corresponding scales, wherein the intermediate feature information is a feature representation of the information to be processed in the target area image;

[0022] Multi-level decoding is performed on the decoding path accordingly, and the first image feature information is produced with the intermediate feature information of the minimum scale, and the image feature information of the previous level and the intermediate feature information produced by the encoding of the same level are used as references step by step to decode the image feature information of higher scales accordingly, and the image feature information is used to represent the contour features of the information to be processed in the target area image in the form of a mask;

[0023] All image feature information is fused to generate mask data of the target contour area in the target area image.

[0024] Optionally, the step of calculating and determining the number of pixels corresponding to the target area contour in each target area image in the urban street view image according to the mask data, and calculating and determining the urban environment characteristic factor according to the number of pixels corresponding to the target area contours includes:

[0025] Based on the mask data corresponding to the green plant area outline, the motor vehicle area outline, the sky area outline, the building area outline, the fence area outline, the pillar area outline and the wall area outline in the urban street view image, the number of pixels corresponding to the green plant area outline, the motor vehicle area outline, the sky area outline, the building area outline, the fence area outline, the pillar area outline and the wall area outline in the urban street view image is calculated and determined;

[0026] Calculate and determine a first ratio between the number of pixels corresponding to the outline of the green plant area and the total number of pixels in the urban street view image, and use the first ratio as the green viewing rate to determine the green viewing rate;

[0027] Calculating and determining a second ratio between the number of pixels corresponding to the motor vehicle area outline and the total number of pixels in the urban street view image, and using the second ratio as a vehicle interference index to determine the vehicle interference index;

[0028] Calculating and determining a third ratio between the number of pixels corresponding to the sky area contour and the total number of pixels in the city street view image, and using the third ratio as a sky visibility index to determine the sky visibility index;

[0029] The sum of the numbers of pixels corresponding to the building area outline, the green plant area outline, the fence area outline, the pillar area outline and the wall area outline is calculated and determined; based on a fourth ratio between the sum and the total number of pixels in the city street view image, the fourth ratio is used as the space enclosure index to determine the space enclosure index.

[0030] Optional, based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 The steps for the fluctuation ratio in pollution include:

[0031] Get the PM of the traffic section where the city street view image is located 2.5 The maximum concentration and PM at a certain time 2.5 concentration;

[0032] Calculate and determine the PM of the traffic section where the urban street view image is located 2.5 The maximum concentration value and the PM at a certain time in the traffic section where the urban street view image is located 2.5 a first difference between the concentrations;

[0033] Based on the first difference and the PM at a certain time of the traffic section where the urban street view image is located. 2.5 The fifth ratio between the concentrations is taken as PM 2.5 The fluctuation ratio of pollution in order to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Fluctuation ratio in pollution.

[0034] Optional, based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 The steps for post-contamination recovery include:

[0035] The traffic section where the city street view image is located is designated as PM 2.5 PM corresponding to the previous day of pollution 2.5 Concentration, PM removal 2.5 PM corresponding to the day after pollution 2.5 Concentration and PM of the traffic section where the urban street view image is located 2.5 Maximum concentration;

[0036] Calculate and determine the PM of the traffic section where the urban street view image is located 2.5 The highest concentration value is consistent with the PM 2.5 PM corresponding to the day after pollution 2.5 a second difference between the concentrations;

[0037] Based on the second difference and the PM 2.5 PM corresponding to the previous day of pollution 2.5 The sixth ratio between the concentrations of 2.5 The recovery ratio after pollution is used to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Recovery ratio after contamination.

[0038] Optional, based on the PM 2.5 Fluctuations in pollution and the PM 2.5 The recovery ratio after pollution is calculated to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 The steps of calculating the elastic coefficient of the road section include:

[0039] Determine the PM corresponding to the traffic section where the city street view image is located 2.5 The fluctuation ratio in pollution and the PM corresponding to the traffic section where the urban street view image is located 2.5 Recovery ratio after contamination;

[0040] Calculate and determine the PM 2.5 The recovery ratio after pollution is related to the PM 2.5 The seventh ratio between the fluctuation ratios in pollution is taken as PM 2.5 The road section elasticity coefficient is used to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Elasticity coefficient of road section.

[0041] Optionally, the vehicle-mounted camera device includes one or more of a driving recorder, a roof camera, a front camera, a rear camera or a side camera; the basic network architecture of the image segmentation model is U 2 net model; the basic network architecture of the atmospheric particulate matter concentration prediction model is a machine learning model, and the machine learning model includes a random forest model.

[0042] A device for predicting urban traffic atmospheric particulate matter concentration provided for another purpose of the present application comprises:

[0043] The data acquisition module is configured to respond to the urban traffic atmospheric particulate matter concentration prediction instruction, obtain multiple city street scene images continuously shot by the vehicle-mounted camera device and their corresponding geographical location information, PM concentration of the traffic section where they are located, and 2.5 Concentration, wherein the geographic location information includes one or more of POI information, normalized vegetation index, and building density;

[0044] a mask data extraction module, configured to use an image segmentation model that has been trained to a convergent state to perform image segmentation on the urban street view image, so as to extract mask data corresponding to the contour of a target area in each target area image in the urban street view image, wherein the target area includes a green plant area, a motor vehicle area, a sky area, a building area, a fence area, a column area, and a wall area;

[0045] An environmental feature determination module is configured to calculate and determine the number of pixels corresponding to the target area contour in each target area image in the urban street view image according to the mask data, and calculate and determine the urban environmental feature factor according to the number of pixels corresponding to the contour of each target area, wherein the urban environmental feature factor includes a green view rate, a vehicle interference index, a sky visibility index, and a space enclosure index;

[0046] The elastic coefficient determination module is configured based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 Fluctuation rates in pollution and PM 2.5 Recovery ratio after pollution, based on the PM 2.5 Fluctuations in pollution and the PM 2.5 The recovery ratio after pollution is calculated to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Section elasticity coefficient;

[0047] PM 2.5 The concentration prediction module is configured based on the geographical location information, urban environment characteristic factors, fluctuation ratio, recovery ratio, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The concentration of the target city is used to construct a sample data set, and the atmospheric particulate matter concentration prediction model trained with the sample data set is used to predict the PM2.5 concentration of the target city corresponding to the traffic section where the street view image is located. 2.5 concentration to complete the prediction of urban traffic atmospheric particulate matter concentration.

[0048] An electronic device provided to meet another purpose of the present application includes a central processing unit and a memory, wherein the central processing unit is used to call and run a computer program stored in the memory to execute the steps of the urban traffic atmospheric particulate matter concentration prediction method described in the present application.

[0049] A computer-readable storage medium is provided to meet another purpose of the present application, which stores a computer program implemented according to the urban traffic atmospheric particulate matter concentration prediction method in the form of computer-readable instructions. When the computer program is called and executed by a computer, the steps included in the corresponding method are executed.

[0050] Compared with the prior art, this application aims at the problems that mobile monitoring in the prior art mainly relies on handheld mobile sensors, which requires high labor costs and insufficient coverage, and the CFD model has high hardware requirements, long simulation time, and high prediction costs. This application includes but is not limited to the following beneficial effects:

[0051] First, this application can greatly expand the monitoring range and temporal resolution. By using image analysis and deep learning technology, it breaks through the spatial limitations of traditional fixed monitoring sites and covers a wider urban area. At the same time, through continuous image data collection, it greatly improves the temporal resolution and can capture PM in a timely manner. 2.5 The dynamic changes of PM concentrations can be analyzed by this comprehensive analysis. 2.5 The accuracy of pollution source tracking has also improved the PM 2.5 Real-time monitoring capability of concentration change trends.

[0052] Secondly, this application uses the geographical location information corresponding to the traffic section in the urban street view image, the urban environment characteristic factors, PM 2.5 Fluctuation rate in pollution, PM 2.5 Recovery rate after pollution, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The concentration of the sample data set is constructed, and the atmospheric particulate matter concentration prediction model trained with the sample data set is used to predict the PM2.5 concentration corresponding to the traffic section where the street view image of the target city is located. 2.5 concentration, which can significantly increase PM 2.5 The prediction accuracy of PM concentration was improved by using a comparative selection strategy to build a prediction model and optimizing parameters. 2.5 The accuracy of concentration prediction: by comprehensively analyzing historical data and real-time data, the model can adapt to a variety of environmental scenarios, enhancing the stability and reliability of the prediction;

[0053] Third, this application can be applied to vehicle-mounted terminal equipment to achieve PM 2.5 The real-time response and decision support capabilities of pollution enable the urban traffic atmospheric particulate matter concentration prediction system to quickly process real-time updated data and immediately provide PM 2.5 The concentration forecast value provides real-time information for city managers, environmental protection workers and the public. Information can be transmitted efficiently and conveniently through vehicle terminals, mobile applications or web platforms, which is conducive to taking corresponding measures quickly to improve air quality.

[0054] Fourth, the PM of this application 2.5 Fluctuation rate in pollution, PM 2.5 Recovery rate after pollution and PM 2.5 The road segment elasticity coefficient can be used to quantify the air quality resilience.2.5 During the concentration process, PM 2.5 Fluctuation rate in pollution, PM 2.5 Recovery rate after pollution and PM 2.5 The road section elasticity coefficient can measure the self-repair ability of air quality under specific conditions, as well as the speed and ability to recover to a better state. 2.5 Fluctuation rate in pollution, PM 2.5 Recovery rate after pollution and PM 2.5 By continuously correcting the elastic coefficient of the road section, the prediction model can better adapt to the changes in environmental factors and ensure the stability and reliability of the prediction results.

[0055] Furthermore, the present application proposes an innovative atmospheric particulate matter concentration tracing and prediction method to address the lag and roughness problems existing in the traditional prediction process of atmospheric particulate matter concentration. This method is based on a deep learning framework, which transforms and fuses high-resolution urban street view images with satellite remote sensing images, thereby extracting the characteristics of the urban built environment in the images. Compared with traditional models built based on social panel statistical data, the prediction model architecture of the present application has a shorter time step and a stronger ability to capture spatial details. Specifically, the present application first uses deep learning technology to fuse high-resolution urban street view images and satellite remote sensing images, uses a deep learning image segmentation model to identify street view images, extracts places or objects in street view images, calculates the pixel coverage of different places or objects, constructs urban environmental characteristic factors, extracts urban environmental characteristic factors that significantly affect atmospheric particulate matter concentrations, and then constructs an atmospheric particulate matter pollution event analysis framework based on urban environmental characteristic factors. Finally, PM is introduced. 2.5 Fluctuation rate in pollution, PM 2.5 Recovery rate after pollution and PM 2.5 Parameters such as road section elasticity coefficient can realize accurate analysis of the changing characteristics of atmospheric particulate pollution in time and space.

[0056] Furthermore, the urban traffic atmospheric particulate matter concentration prediction method proposed in this application can effectively solve the lag and roughness problems existing in traditional prediction methods; the urban traffic atmospheric particulate matter concentration prediction method proposed in this application has broad application prospects in the field of intelligent management of atmospheric particulate matter pollution in different scenarios, and is expected to provide effective technical support for improving urban air quality and pollution prevention and control. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0058] Figure 1A schematic diagram of the process of predicting the concentration of particulate matter in urban traffic atmosphere in an embodiment of the present application;

[0059] Figure 2 This is an interface display diagram of the urban traffic atmospheric particulate matter concentration prediction system in the embodiment of the present application;

[0060] Figure 3 In the embodiment of this application, 2 An exemplary network architecture of the net model;

[0061] Figure 4 This is a functional block diagram of a device for predicting the concentration of particulate matter in urban traffic atmosphere in an embodiment of the present application;

[0062] Figure 5 It is a schematic diagram of the structure of the computer device in the embodiment of the present application. DETAILED DESCRIPTION

[0063] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be interpreted as limiting the present application.

[0064] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0065] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with those in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless specifically defined as here.

[0066] It will be understood by those skilled in the art that the "client", "terminal" and "terminal device" used herein include both devices with wireless signal receivers, which are devices with only wireless signal receivers without transmission capabilities, and devices with receiving and transmitting hardware, which are devices with receiving and transmitting hardware capable of two-way communication on a two-way communication link. Such devices may include: cellular or other communication devices such as personal computers, tablet computers, which have single-line displays or multi-line displays or cellular or other communication devices without multi-line displays; PCS (Personal Communications Service, personal communication system), which can combine voice, data processing, fax and / or data communication capabilities; PDA (Personal Digital Assistant, personal digital assistant), which may include a radio frequency receiver, pager, Internet / intranet access, web browser, notepad, calendar and / or GPS (Global Positioning System, global positioning system) receiver; conventional laptop and / or palmtop computers or other devices, which have and / or include a conventional laptop and / or palmtop computer or other device with and / or including a radio frequency receiver. The "client", "terminal" and "terminal device" used herein may be portable, transportable, installed in a vehicle (air, sea and / or land), or suitable for and / or configured to run locally, and / or in a distributed form, at any other location on the earth and / or in space. The "client", "terminal" and "terminal device" used herein may also be a communication terminal, an Internet terminal, a music / video playing terminal, for example, a PDA, a MID (Mobile Internet Device) and / or a mobile phone with a music / video playing function, or a smart TV, a set-top box and other devices.

[0067] The hardware referred to by the names such as "server", "client", and "service node" in this application is essentially an electronic device with the equivalent capabilities of a personal computer. It is a hardware device with the necessary components revealed by the von Neumann principle, such as a central processing unit (including an arithmetic unit and a controller), a memory, an input device, and an output device. The computer program is stored in its memory, and the central processing unit calls the program stored in the external memory into the internal memory for execution, executes the instructions in the program, and interacts with the input and output devices to complete specific functions.

[0068] It should be pointed out that the concept of "server" referred to in this application can also be extended to the case of server clusters. According to the network deployment principle understood by those skilled in the art, the servers should be logically divided. In physical space, these servers can be independent of each other but can be called through interfaces, or integrated into a physical computer or a set of computer clusters. Those skilled in the art should understand this flexibility, and should not use it to restrict the implementation of the network deployment method of this application.

[0069] Unless expressly specified, one or more technical features of the present application can be deployed on a server for implementation and accessed by a client through a remote call to obtain an online service interface provided by the server, or can be directly deployed and run on a client for access.

[0070] The neural network models referenced or may be referenced in this application, unless expressly specified, can be deployed on a remote server and remotely called on the client, or can be deployed and directly called on a client with sufficient device capabilities. In some embodiments, when it runs on the client, its corresponding intelligence can be obtained through transfer learning to reduce the requirements for the client's hardware operating resources and avoid excessive occupation of the client's hardware operating resources.

[0071] Unless explicitly specified, the various data involved in this application can be stored remotely on a server or on a local terminal device, as long as it is suitable for being called by the technical solution of this application.

[0072] Those skilled in the art should be aware that, although the various methods of the present application are described based on the same concept and thus present commonality to each other, unless otherwise specified, these methods can be independently executed. Similarly, for each embodiment disclosed in the present application, they are all proposed based on the same inventive concept, therefore, concepts with the same expression, and concepts that are appropriately changed for convenience despite different expressions, should be understood as equivalent.

[0073] Unless the mutually exclusive relationship between the embodiments to be disclosed in this application is explicitly stated, the relevant technical features involved in each embodiment can be cross-combined to flexibly construct a new embodiment, as long as such combination does not deviate from the creative spirit of this application and can meet the needs of the prior art or solve certain deficiencies in the prior art. Those skilled in the art should be aware of this flexibility.

[0074] Based on the above exemplary scenarios, please refer to Figure 1 as well as Figure 2 In one embodiment, the method for predicting the concentration of particulate matter in urban traffic atmosphere of the present application includes:

[0075] Step S10: respond to the urban traffic atmospheric particulate matter concentration prediction instruction, obtain multiple city street scene images taken continuously by the vehicle-mounted camera device and their corresponding geographical location information, PM concentration of the traffic section where they are located, and 2.5 Concentration, wherein the geographic location information includes one or more of POI information, normalized vegetation index, and building density;

[0076] The urban traffic atmospheric particulate matter concentration prediction system in the vehicle terminal device, mobile terminal device or web platform can respond to the urban traffic atmospheric particulate matter concentration prediction instruction, obtain multiple city street scene images taken continuously by the vehicle camera device and their corresponding geographical location information, PM2.5 concentration of the traffic section, and calculate the concentration of the urban traffic atmospheric particulate matter. 2.5 concentration, wherein the geographic location information includes one or more of POI information, normalized vegetation index, and building density; the vehicle-mounted camera equipment includes one or more of a driving recorder, a roof camera, a front camera, a rear camera, or a side-view camera;

[0077] Specifically, the type and source of the city street view image depend on the actual application scenario. For example, in the application scenario of urban traffic atmospheric particulate matter concentration prediction, the city street view image can be a static picture specified by the user, or it can be a city street view image frame in the city street view video stream submitted by a camera device in a moving vehicle, an unmanned vehicle or an autonomous driving vehicle (such as a driving recorder, a roof camera, a front camera, a rear camera or a side view camera, etc.) to a vehicle-mounted terminal device. Depending on the specific application scenario, the city street view image can be determined as needed.

[0078] In some embodiments, the POI information provides the function and purpose of the traffic section where the urban street view image is located, such as business, education, entertainment, etc. This information is helpful for analyzing the economic and social activities of the region to predict PM. 2.5 The normalized difference vegetation index (NDVI) is an indicator of vegetation coverage and health status, which can reflect the ecological environment of the region. A high NDVI value usually indicates lush vegetation and good environmental quality, which is consistent with the PM in the traffic section of the urban street view image. 2.5 The building density reflects the utilization degree of the urban space of the traffic section where the urban street view image is located, which affects the population density and traffic flow of the area. High-density areas are often accompanied by PM 2.5 Higher concentration.

[0079] In some embodiments, the real-time positioning capability of the vehicle can be used to send data requests to the map service provider. The request should include specific spatial scope and attribute requirements. After accessing the map service provider, the geographic location information returned by the map service provider is received, including POI information, NVDI, building density, etc., to ensure that the collected urban street view images are strictly synchronized with the geographic location information. The urban traffic atmospheric particulate matter concentration prediction system can automatically record the capture time and location of the image, so that the image analysis results are closely coupled with the actual environmental conditions. Through time series analysis, the changes in environmental characteristics on the driving path are combined with the time dimension to analyze PM 2.5 Correlation between concentration and urban environmental characteristic factors.

[0080] Step S20, using the image segmentation model that has been trained to a convergent state to perform image segmentation on the urban street view image, so as to extract mask data corresponding to the target area contour in each target area image in the urban street view image, wherein the target area includes a green plant area, a motor vehicle area, a sky area, a building area, a fence area, a column area, and a wall area;

[0081] Obtain multiple city street scene images taken continuously by the vehicle-mounted camera device and their corresponding geographical location information, PM of the traffic section where they are located 2.5 After the concentration, the image segmentation model that has been trained to a convergent state is used to perform image segmentation on the urban street view image to extract mask data corresponding to the target area contour in each target area image in the urban street view image, wherein the target area includes a green plant area, a motor vehicle area, a sky area, a building area, a fence area, a column area and a wall area; the basic network architecture of the image segmentation model can be U 2 net model.

[0082] Furthermore, the step of using the image segmentation model that has been trained to a convergent state to segment the city street view image to extract mask data corresponding to the target area contour in each target area image in the city street view image includes:

[0083] Step S201, performing multi-level encoding on the target area image, and correspondingly generating intermediate feature information of corresponding scales, wherein the intermediate feature information is a feature representation of the information to be processed in the target area image;

[0084] See also Figure 3 According to the six encoders (En_1 to En_6) of the side branch path, the standard original image corresponding to the urban street view image, that is, the original image adapted to U 2net model needs to meet the specifications of the input image, and the city street view image is cropped into a city street view image of the specified specifications, and the original image of this specification is encoded step by step. The first-level encoder at the top layer extracts the intermediate feature information corresponding to the first scale from the original image of the specification, and then passes it to the encoder at the next level to extract the intermediate feature information corresponding to the second scale. By analogy, the six-level encoder extracts the corresponding intermediate feature information. It can be seen that after the original image of the specification is encoded step by step through the encoding path, six intermediate feature information of corresponding spatial resolutions can be obtained.

[0085] It can be understood that the intermediate feature information of each scale is a representation obtained after deep semantic understanding of the standard image at the corresponding scale, and is information extracted from the contour features of the human body in the urban street view image. 2 This capability of the net model is well known to those skilled in the art. As long as it is trained to convergence with a sufficient number of training samples, its encoding path can have the ability to capture the deep semantic understanding of the information to be processed in the target area image.

[0086] Step S202: perform multi-level decoding on the decoding path accordingly, generate first image feature information with the smallest-scale intermediate feature information, and decode higher-scale image feature information correspondingly with reference to the upper-level image feature information and the intermediate feature information generated by the same-level encoding, wherein the image feature information is used to represent the contour features of the information to be processed in the target area image in the form of a mask;

[0087] See also Figure 3 , in the decoding path, in the five decoders (De_1 to De_5) of the right branch path, starting from the bottom layer, each decoder stage takes the upsampled feature map from the previous stage and the cascade of the feature map from its symmetric encoder stage as input, decodes and outputs feature information. The output results of the last encoder stage (En_6) and the decoder stages of each level are subjected to 1*1 convolution kernel transformation, Sigmoid activation function and upsampling operation to extract six image feature information of the same specifications as the original image. The image feature information can characterize the green plant area contour information, motor vehicle area contour information, sky area contour information, building area contour information, fence area contour information, pillar area contour information or wall area in the urban street scene image, and the contour features are represented in the form of masks, which are essentially mask image data (Mask1 to Mask6).

[0088] Step S203: Fusing all image feature information to generate mask data of the target contour area in the target area image.

[0089] At this point, based on the disclosure of the structure and principles of the image segmentation model herein, it can be understood that the target area image of the present application, after being encoded and decoded step by step by the image segmentation model, obtains multiple image feature information (mask image data: Mask1 to Mask6), and all image feature information is fused to generate mask data of the target contour area in the target area image.

[0090] Step S30, calculating and determining the number of pixels corresponding to the target area contour in each target area image in the urban street view image according to the mask data, and calculating and determining the urban environment characteristic factors according to the number of pixels corresponding to the target area contour, wherein the urban environment characteristic factors include green view rate, vehicle interference index, sky visibility index and space enclosure index;

[0091] The image segmentation model that has been trained to a convergent state is used to perform image segmentation on the city street view image to extract mask data corresponding to the target area contour in each target area image in the city street view image, and then the number of pixels corresponding to the target area contour in each target area image in the city street view image is calculated and determined according to the mask data. The urban environment characteristic factors are calculated and determined according to the number of pixels corresponding to each target area contour, and the urban environment characteristic factors include green viewing rate, vehicle interference index, sky visibility index and space enclosure index; it is not difficult to understand that when the mask data corresponding to the target area contour in each target area image in the city street view image is determined, the number of pixels corresponding to the target area contour in each target area image in the city street view image is also determined, which will not be elaborated here.

[0092] In some embodiments, the step of calculating and determining the number of pixels corresponding to the target area contour in each target area image in the urban street view image according to the mask data, and calculating and determining the urban environment characteristic factor according to the number of pixels corresponding to each target area contour includes:

[0093] Step S301, based on mask data corresponding to the green plant area outline, motor vehicle area outline, sky area outline, building area outline, fence area outline, pillar area outline and wall area outline in the urban street view image, calculate and determine the number of pixels corresponding to the green plant area outline, motor vehicle area outline, sky area outline, building area outline, fence area outline, pillar area outline and wall area outline in the urban street view image;

[0094] Step S302, calculating and determining a first ratio between the number of pixels corresponding to the green plant area outline and the total number of pixels in the urban street view image, and taking the first ratio as the green viewing rate to determine the green viewing rate;

[0095] Specifically, the green view index (GVI) is used to quantify three-dimensional street greening, evaluate the visibility of urban greening, and reflect people's visual perception of surrounding green spaces in three dimensions. It is defined as the green pixels in the urban street view image divided by the total pixels. The calculation formula of the green view index is as follows:

[0096]

[0097] Where Area g Area is the number of pixels corresponding to the outline of the green area in the urban street view image. i is the total number of pixels in the urban street view image, and GVI represents the green view rate in the urban street view image.

[0098] Step S303, calculating and determining a second ratio between the number of pixels corresponding to the motor vehicle area outline and the total number of pixels in the city street view image, and using the second ratio as a vehicle interference index to determine the vehicle interference index;

[0099] Specifically, the vehicle interference index (VII) is an indicator to measure the concentration of motor vehicles on urban roads. VII is a reflection of the amount of travel by urban residents and has a great impact on PM. 2.5 The emission and concentration of the vehicle interference index are calculated as follows:

[0100]

[0101] Where VII represents the vehicle interference index, C n is the number of pixels corresponding to the outline of the motor vehicle area in the nth city street view image, A n is the total number of pixels in the street view image of the nth city.

[0102] Step S304, calculating and determining a third ratio between the number of pixels corresponding to the sky area outline and the total number of pixels in the city street view image, and using the third ratio as a sky visibility index to determine the sky visibility index;

[0103] Specifically, the sky visibility index (SVI) refers to the degree of streetscape permeability, which has a significant impact on PM 2.5 The spatial circulation has an important influence, and its calculation formula is as follows:

[0104]

[0105] Where SVI is the sky visibility index, V n is the number of pixels corresponding to the sky area contour in the nth city street view image, A n is the total number of pixels in the street view image of the nth city.

[0106] Step S305, calculate and determine the sum of the numbers of pixels corresponding to the building area outline, the green plant area outline, the fence area outline, the pillar area outline and the wall area outline, and based on a fourth ratio between the sum and the total number of pixels in the city street view image, use the fourth ratio as the space enclosure index to determine the space enclosure index.

[0107] Specifically, the space enclosure index (SEI) refers to the degree of enclosure of the overall space on both sides of the street. The degree of maintenance determines the PM 2.5 The diffusion efficiency of PM 2.5 It is easier to be trapped in the street space. SEI is determined by the proportion of facade factors (such as buildings, plants and walls), and its calculation formula is as follows:

[0108]

[0109] In the formula, SEI represents the space enclosure index, B n Represents the number of pixels corresponding to the building area outline, G n Indicates the number of pixels corresponding to the outline of the green plant area, Q n Indicates the number of pixels corresponding to the outline of the fence area, P n Indicates the number of pixels corresponding to the outline of the column area, F n Indicates the number of pixels corresponding to the outline of the wall area.

[0110] Step S40: Based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 Fluctuation rates in pollution and PM 2.5 Recovery ratio after pollution, based on the PM 2.5 Fluctuations in pollution and the PM 2.5 The recovery ratio after pollution is calculated to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Section elasticity coefficient;

[0111] According to the mask data, the number of pixels corresponding to the target area contour in each target area image in the urban street view image is calculated and determined. After the urban environment characteristic factor is calculated and determined according to the number of pixels corresponding to the target area contour, based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 Fluctuation rates in pollution and PM 2.5 Recovery ratio after pollution, based on the PM 2.5 Fluctuations in pollution and the PM2.5 The recovery ratio after pollution is calculated to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Section elasticity coefficient;

[0112] Further, based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 The steps for the fluctuation ratio in pollution include:

[0113] Step S401: Obtain the PM of the traffic section where the city street view image is located. 2.5 The maximum concentration and PM at a certain time 2.5 concentration;

[0114] Step S402: Calculate and determine the PM of the traffic section where the city street view image is located. 2.5 The maximum concentration value and the PM at a certain time in the traffic section where the urban street view image is located 2.5 the first difference between the concentrations;

[0115] Step S403: based on the first difference and the PM value of the traffic section where the urban street view image is located at a certain time, 2.5 The fifth ratio between the concentrations is taken as PM 2.5 The fluctuation ratio of pollution in order to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Fluctuation ratio in pollution.

[0116] Specifically, the PM corresponding to the traffic section where the urban street view image is located 2.5 The fluctuation ratio in pollution is calculated as:

[0117]

[0118] Among them, FR i Indicates the i-th traffic section in the city street view image in PM 2.5 Fluctuation ratios in pollution; Indicates the i-th traffic section in the city street view image at t m PM of time 2.5 The highest concentration, represents the PM of the i-th traffic section at time t in the city street view image 2.5 Concentration, t m Indicates PM 2.5 The time when pollution reaches its peak.

[0119] Further, based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located.2.5 The steps for post-contamination recovery include:

[0120] Step S4001: Obtain the traffic section where the urban street view image is located as PM 2.5 PM corresponding to the previous day of pollution 2.5 Concentration, PM removal 2.5 PM corresponding to the day after pollution 2.5 Concentration and PM of the traffic section where the urban street view image is located 2.5 Maximum concentration;

[0121] Step S4002: Calculate and determine the PM of the traffic section where the city street view image is located. 2.5 The highest concentration value is consistent with the PM 2.5 PM corresponding to the day after pollution 2.5 a second difference between the concentrations;

[0122] Step S4003: based on the second difference and the PM 2.5 PM corresponding to the previous day of pollution 2.5 The sixth ratio between the concentrations of 2.5 The recovery ratio after pollution is used to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Recovery ratio after contamination.

[0123] Specifically, the PM corresponding to the traffic section where the urban street view image is located 2.5 The recovery ratio after pollution is calculated as follows:

[0124]

[0125] Among them, RR i Indicates the i-th traffic section in the city street view image in PM 2.5 Recovery ratio after contamination; Indicates the i-th traffic section in the city street view image at t m PM of time 2.5 Maximum concentration; Indicates the i-th traffic section in the city street view image at t 1 PM corresponding to the time 2.5 concentration; Indicates the i-th traffic section in the city street view image at t 2 PM corresponding to the time 2.5 Concentration; t 1 Indicates the PM 2.5 The day before the pollution; t 2 Indicates the release of PM 2.5 The day after the pollution; tm Indicates PM 2.5 The time when pollution reaches its peak.

[0126] Furthermore, based on the PM 2.5 Fluctuations in pollution and the PM 2.5 The recovery ratio after pollution is calculated to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 The steps of calculating the elastic coefficient of the road section include:

[0127] Step S1100: Determine the PM corresponding to the traffic section where the city street view image is located. 2.5 The fluctuation ratio in pollution and the PM corresponding to the traffic section where the urban street view image is located 2.5 Recovery ratio after contamination;

[0128] Step S1200: Calculate and determine the PM 2.5 The recovery ratio after pollution is related to the PM 2.5 The seventh ratio between the fluctuation ratios in pollution is taken as PM 2.5 The road section elasticity coefficient is used to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Elasticity coefficient of road section.

[0129] Specifically, the PM corresponding to the traffic section where the urban street view image is located 2.5 The calculation formula of the road section elastic coefficient is expressed as:

[0130]

[0131] Among them, RR i Indicates the i-th traffic section in the city street view image in PM 2.5 Recovery ratio after contamination; FR i Indicates the i-th traffic section in the city street view image in PM 2.5 Fluctuation ratio in pollution; CRE i，t represents the PM corresponding to the i-th traffic section at time t in the city street view image 2.5 Elasticity coefficient of road section.

[0132] Step S50: Based on the geographic location information, urban environment characteristic factors, fluctuation ratio, recovery ratio, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The concentration of the target city is used to construct a sample data set, and the atmospheric particulate matter concentration prediction model trained with the sample data set is used to predict the PM2.5 concentration of the target city corresponding to the traffic section where the street view image is located. 2.5 concentration to complete the prediction of urban traffic atmospheric particulate matter concentration.

[0133] Calculate and determine the PM corresponding to the traffic section where the urban street view image is located 2.5 After the road section elasticity coefficient, based on the geographical location information, urban environment characteristic factors, fluctuation ratio, recovery ratio, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The concentration of the target city is used to construct a sample data set, and the atmospheric particulate matter concentration prediction model trained with the sample data set is used to predict the PM2.5 concentration of the target city corresponding to the traffic section where the street view image is located. 2.5 Concentration is used to complete the prediction of urban traffic atmospheric particulate matter concentration, wherein the basic network architecture of the atmospheric particulate matter concentration prediction model is a machine learning model, and the machine learning model includes a random forest model, a generalized additive model (Generalized Additive Model, GAM) or a support vector machine (Support Vector Machine, SVM), etc. The atmospheric particulate matter concentration prediction model of the present application can be selected from the above models.

[0134] In some embodiments, the present application adopts a comparative selection strategy to construct an atmospheric particulate matter concentration prediction model. This strategy involves pairwise comparison of all potential influencing factors, combining the above-mentioned geographic location information, urban environmental characteristic factors, fluctuation ratio, recovery ratio, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The most predictive feature combination is gradually screened out by combining the concentration. Through iterative optimization, the simplest and most efficient feature set is determined to improve the explanatory power and prediction accuracy of the model. Then, multiple machine learning models are constructed, and the screened feature sets are put into the machine learning models. By comparing the determination coefficient (R 2 ), mean absolute percentage error (MAPE), root mean square error (RMSE), mean absolute error (MAE), complete the performance evaluation of the machine learning algorithm, and find the machine learning algorithm with the best prediction performance. RMSE is used to evaluate the average difference between the predicted value and the actual value. The smaller its value, the smaller the difference between the predicted value and the actual value of the model, and the better the model performance; MAE is used to evaluate the average difference between the predicted value and the actual value. It can better handle outliers. The smaller its value, the smaller the difference between the predicted value and the actual value of the model, and the better the model performance; R 2 The value range of is 0 to 1. The closer it is to 1, the better the model explains the variability of the dependent variable and the better the model performance. When R is equal to 1, it means that the model can perfectly explain the variability of the dependent variable and the predicted value is completely consistent with the actual value. The performance of machine learning methods is measured by the coefficient of determination (R 2), mean absolute percentage error (MAPE), root mean square error (RMSE), and mean absolute error (MAE) are calculated according to the following formulas:

[0135]

[0136] In the formula, n is the number of samples, y is i is the true value of each sample, is the mean of the sample data, is the predicted value of the sample. The smaller the RMSE value, MAE value and MAPE value, the higher the accuracy of the model in describing the data. 2 The value of R is between [0,1], 2 The closer it is to 1, the stronger the explanatory power of the predictor variable for the predicted object and the better the model fits the data.

[0137] In some embodiments, the basic network architecture of the atmospheric particulate matter concentration prediction model of the present application is a random forest model, which does not constitute a limitation to the present application, and the geographical location information, urban environment characteristic factors, fluctuation ratio, recovery ratio, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The sample data set is constructed by using the random forest model to train the model until the model is trained to a convergence state or reaches a certain number of iterations. After the atmospheric particulate matter concentration prediction model is trained to a convergence state, it can be put into production and used to predict the PM corresponding to the traffic section where the street view image of the target city is located. 2.5 concentration.

[0138] As can be seen from the above embodiments, compared with the prior art, the present application aims at the problems that mobile monitoring in the prior art mainly relies on handheld mobile sensors, which requires high labor costs and insufficient coverage, and the CFD model has high hardware requirements, long simulation time, and high prediction costs. The present application includes but is not limited to the following beneficial effects:

[0139] First, this application can greatly expand the monitoring range and temporal resolution. By using image analysis and deep learning technology, it breaks through the spatial limitations of traditional fixed monitoring sites and covers a wider urban area. At the same time, through continuous image data collection, it greatly improves the temporal resolution and can capture PM in a timely manner. 2.5 The dynamic changes of PM concentrations can be analyzed by this comprehensive analysis. 2.5 The accuracy of pollution source tracking has also improved the PM 2.5 Real-time monitoring capability of concentration change trends.

[0140] Secondly, this application uses the geographical location information corresponding to the traffic section in the urban street view image, the urban environment characteristic factors, PM2.5 Fluctuation rate in pollution, PM 2.5 Recovery rate after pollution, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The concentration of the sample data set is constructed, and the atmospheric particulate matter concentration prediction model trained with the sample data set is used to predict the PM2.5 concentration corresponding to the traffic section where the street view image of the target city is located. 2.5 concentration, which can significantly increase PM 2.5 The prediction accuracy of PM concentration was improved by using a comparative selection strategy to build a prediction model and optimizing parameters. 2.5 The accuracy of concentration prediction: by comprehensively analyzing historical data and real-time data, the model can adapt to a variety of environmental scenarios, enhancing the stability and reliability of the prediction;

[0141] Third, this application can be applied to vehicle-mounted terminal equipment to achieve PM 2.5 The real-time response and decision support capabilities of pollution enable the urban traffic atmospheric particulate matter concentration prediction system to quickly process real-time updated data and immediately provide PM 2.5 The concentration forecast value provides real-time information for city managers, environmental protection workers and the public. Information can be transmitted efficiently and conveniently through vehicle terminals, mobile applications or web platforms, which is conducive to taking corresponding measures quickly to improve air quality.

[0142] Fourth, the PM of this application 2.5 Fluctuation rate in pollution, PM 2.5 Recovery rate after pollution and PM 2.5 The road segment elasticity coefficient can be used to quantify the air quality resilience. 2.5 During the concentration process, PM 2.5 Fluctuation rate in pollution, PM 2.5 Recovery rate after pollution and PM 2.5 The road section elasticity coefficient can measure the self-repair ability of air quality under specific conditions, as well as the speed and ability to recover to a better state. 2.5 Fluctuation rate in pollution, PM 2.5 Recovery rate after pollution and PM 2.5 The continuous correction of the elastic coefficient of the road section enables the prediction model to better adapt to the changes in environmental factors and ensure the stability and reliability of the prediction results.

[0143] Furthermore, the present application proposes an innovative atmospheric particulate matter concentration tracing and prediction method to address the lag and roughness problems existing in the traditional prediction process of atmospheric particulate matter concentration. This method is based on a deep learning framework, which transforms and fuses high-resolution urban street view images with satellite remote sensing images, thereby extracting the characteristics of the urban built environment in the images. Compared with traditional models built based on social panel statistical data, the prediction model architecture of the present application has a shorter time step and a stronger ability to capture spatial details. Specifically, the present application first uses deep learning technology to fuse high-resolution urban street view images and satellite remote sensing images, uses a deep learning image segmentation model to identify street view images, extracts places or objects in street view images, calculates the pixel coverage of different places or objects, constructs urban environmental characteristic factors, extracts urban environmental characteristic factors that significantly affect atmospheric particulate matter concentrations, and then constructs an atmospheric particulate matter pollution event analysis framework based on urban environmental characteristic factors. Finally, PM is introduced. 2.5 Fluctuation rate in pollution, PM 2.5 Recovery rate after pollution and PM 2.5 Parameters such as road section elasticity coefficient can realize accurate analysis of the changing characteristics of atmospheric particulate pollution in time and space.

[0144] Furthermore, the urban traffic atmospheric particulate matter concentration prediction method proposed in this application can effectively solve the lag and roughness problems existing in traditional prediction methods; the urban traffic atmospheric particulate matter concentration prediction method proposed in this application has broad application prospects in the field of intelligent management of atmospheric particulate matter pollution in different scenarios, and is expected to provide effective technical support for improving urban air quality and pollution prevention and control.

[0145] See also Figure 4 , adapted to one of the purposes of the present application, a device for predicting the concentration of particulate matter in urban traffic atmosphere is provided, comprising a data acquisition module 1100, a mask data extraction module 1200, an environmental feature determination module 1300, an elastic coefficient determination module 1400 and a PM 2.5 The concentration prediction module 1500 includes a data acquisition module 1100 configured to respond to the urban traffic atmospheric particulate matter concentration prediction instruction, acquire a plurality of urban street scene images continuously shot by the vehicle-mounted camera device, and their corresponding geographical location information, PM concentration of the traffic section, and the like. 2.5concentration, wherein the geographic location information includes one or any multiple of POI information, normalized vegetation index, and building density; a mask data extraction module 1200 is configured to use an image segmentation model that has been trained to a convergence state to perform image segmentation on the urban street view image to extract mask data corresponding to the target area contour in each target area image in the urban street view image, wherein the target area includes a green plant area, a motor vehicle area, a sky area, a building area, a fence area, a column area, and a wall area; an environmental feature determination module 1300 is configured to calculate and determine the number of pixels corresponding to the target area contour in each target area image in the urban street view image according to the mask data, and calculate and determine the urban environment feature factor according to the number of pixels corresponding to the contour of each target area, wherein the urban environment feature factor includes a green view rate, a vehicle interference index, a sky visibility index, and a space enclosure index; an elasticity coefficient determination module 1400 is configured to calculate and determine the urban environment feature factor based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 Fluctuation rates in pollution and PM 2.5 Recovery ratio after pollution, based on the PM 2.5 Fluctuations in pollution and the PM 2.5 The recovery ratio after pollution is calculated to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Section elastic coefficient; PM 2.5 The concentration prediction module 1500 is configured to be based on the geographical location information, urban environment characteristic factors, fluctuation ratio, recovery ratio, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The concentration of the target city is used to construct a sample data set, and the atmospheric particulate matter concentration prediction model trained with the sample data set is used to predict the PM2.5 concentration of the target city corresponding to the traffic section where the street view image is located. 2.5 concentration to complete the prediction of urban traffic atmospheric particulate matter concentration.

[0146] Based on any embodiment of this application, please refer to Figure 5 Another embodiment of the present application further provides an electronic device, which can be implemented by a computer device, such as Figure 5As shown, a schematic diagram of the internal structure of a computer device. The computer device includes a processor, a computer-readable storage medium, a memory, and a network interface connected via a system bus. Among them, the computer-readable storage medium of the computer device stores an operating system, a database, and computer-readable instructions. The database may store a control information sequence. When the computer-readable instructions are executed by the processor, the processor may implement a method for predicting the concentration of particulate matter in the atmosphere of urban traffic. The processor of the computer device is used to provide computing and control capabilities to support the operation of the entire computer device. The memory of the computer device may store computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor may execute the method for predicting the concentration of particulate matter in the atmosphere of urban traffic of the present application. The network interface of the computer device is used to connect and communicate with a terminal. Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0147] In this embodiment, the processor is used to execute Figure 4 The memory stores the program code and various data required to execute the above modules or submodules. The network interface is used to transmit data between user terminals or servers. The memory in this embodiment stores the program code and data required to execute all modules / submodules in the urban traffic atmospheric particulate matter concentration prediction device of this application, and the server can call the program code and data of the server to execute the functions of all submodules.

[0148] The present application also provides a storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the urban traffic atmospheric particulate matter concentration prediction method described in any embodiment of the present application.

[0149] The present application also provides a computer program product, including a computer program / instruction, which, when executed by one or more processors, implements the steps of the method for predicting urban traffic atmospheric particulate matter concentration described in any embodiment of the present application.

[0150] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments of the present application can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a computer-readable storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0151] The above description is only a partial implementation method of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

[0152] To sum up, the urban traffic atmospheric particulate matter concentration prediction method proposed in this application can effectively solve the lag and roughness problems existing in traditional prediction methods; the urban traffic atmospheric particulate matter concentration prediction method proposed in this application has broad application prospects in the field of intelligent management of atmospheric particulate matter pollution in different scenarios, and is expected to provide effective technical support for improving urban air quality and pollution prevention and control.

Claims

1. A method for predicting urban traffic atmospheric particulate matter concentration, characterized in that: include: In response to the urban traffic atmospheric particulate matter concentration prediction command, multiple city street scene images taken continuously by the vehicle-mounted camera equipment and their corresponding geographical location information, PM concentration of the traffic section where they are located are obtained. 2.5 Concentration, wherein the geographic location information includes one or more of POI information, normalized vegetation index, and building density; Performing image segmentation on the urban street view image using an image segmentation model that has been trained to a convergent state, so as to extract mask data corresponding to the contour of a target area in each target area image in the urban street view image, wherein the target area includes a green plant area, a motor vehicle area, a sky area, a building area, a fence area, a column area, and a wall area; Calculate and determine the number of pixels corresponding to the target area contour in each target area image in the urban street view image according to the mask data, and calculate and determine the urban environment characteristic factor according to the number of pixels corresponding to the target area contour, wherein the urban environment characteristic factor includes a green view rate, a vehicle interference index, a sky visibility index, and a space enclosure index; Based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 Fluctuation rates in pollution and PM 2.5 Recovery ratio after pollution, based on the PM 2.5 Fluctuations in pollution and the PM 2.5 The recovery ratio after pollution is calculated to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Section elasticity coefficient; Based on the geographical location information, urban environment characteristic factors, fluctuation ratio, recovery ratio, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The concentration of the target city is used to construct a sample data set, and the atmospheric particulate matter concentration prediction model trained with the sample data set is used to predict the PM2.5 concentration of the target city corresponding to the traffic section where the street view image is located. 2.5 concentration to complete the prediction of urban traffic atmospheric particulate matter concentration.

2. The method for predicting urban traffic atmospheric particulate matter concentration according to claim 1, characterized in that: The step of using an image segmentation model that has been trained to a convergent state to segment the city street view image to extract mask data corresponding to the target area contour in each target area image in the city street view image comprises: Performing multi-level encoding on the target area image and correspondingly generating intermediate feature information of corresponding scales, wherein the intermediate feature information is a feature representation of the information to be processed in the target area image; Multi-level decoding is performed on the decoding path accordingly, and the first image feature information is produced with the intermediate feature information of the minimum scale, and the image feature information of the previous level and the intermediate feature information produced by the encoding of the same level are used as references step by step to decode the image feature information of higher scales accordingly, and the image feature information is used to represent the contour features of the information to be processed in the target area image in the form of a mask; All image feature information is fused to generate mask data of the target contour area in the target area image.

3. The method for predicting urban traffic atmospheric particulate matter concentration according to claim 1, characterized in that: The steps of calculating and determining the number of pixels corresponding to the contour of the target area in each target area image in the urban street view image according to the mask data, and calculating and determining the urban environment characteristic factor according to the number of pixels corresponding to the contour of each target area, include: Based on the mask data corresponding to the green plant area outline, the motor vehicle area outline, the sky area outline, the building area outline, the fence area outline, the pillar area outline and the wall area outline in the urban street view image, the number of pixels corresponding to the green plant area outline, the motor vehicle area outline, the sky area outline, the building area outline, the fence area outline, the pillar area outline and the wall area outline in the urban street view image is calculated and determined; Calculate and determine a first ratio between the number of pixels corresponding to the outline of the green plant area and the total number of pixels in the urban street view image, and use the first ratio as the green viewing rate to determine the green viewing rate; Calculating and determining a second ratio between the number of pixels corresponding to the motor vehicle area outline and the total number of pixels in the urban street view image, and using the second ratio as a vehicle interference index to determine the vehicle interference index; Calculating and determining a third ratio between the number of pixels corresponding to the sky area contour and the total number of pixels in the city street view image, and using the third ratio as a sky visibility index to determine the sky visibility index; The sum of the numbers of pixels corresponding to the building area outline, the green plant area outline, the fence area outline, the pillar area outline and the wall area outline is calculated and determined; based on a fourth ratio between the sum and the total number of pixels in the city street view image, the fourth ratio is used as the space enclosure index to determine the space enclosure index.

4. The method for predicting urban traffic atmospheric particulate matter concentration according to claim 1, characterized in that: Based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 The steps for the fluctuation ratio in pollution include: Get the PM of the traffic section where the city street view image is located 2.5 The maximum concentration and PM at a certain time 2.5 concentration; Calculate and determine the PM of the traffic section where the urban street view image is located 2.5 The maximum concentration value and the PM at a certain time in the traffic section where the urban street view image is located 2.5 the first difference between the concentrations; Based on the first difference and the PM at a certain time of the traffic section where the urban street view image is located. 2.5 The fifth ratio between the concentrations is taken as PM 2.5 The fluctuation ratio of pollution in order to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Fluctuation ratio in pollution.

5. The method for predicting urban traffic atmospheric particulate matter concentration according to claim 1, characterized in that: Based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 The steps for post-contamination recovery include: The traffic section where the city street view image is located is designated as PM 2.5 PM corresponding to the previous day of pollution 2.5 Concentration, PM removal 2.5 PM corresponding to the day after pollution 2.5 Concentration and PM of the traffic section where the urban street view image is located 2.5 Maximum concentration; Calculate and determine the PM of the traffic section where the urban street view image is located 2.5 The highest concentration value is consistent with the PM 2.5 PM corresponding to the day after pollution 2.5 a second difference between the concentrations; Based on the second difference and the PM 2.5 PM corresponding to the previous day of pollution 2.5 The sixth ratio between the concentrations of 2.5 The recovery ratio after pollution is used to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Recovery ratio after contamination.

6. The method for predicting urban traffic atmospheric particulate matter concentration according to claim 1, characterized in that: Based on the PM 2.5 Fluctuations in pollution and the PM 2.5 The recovery ratio after pollution is calculated to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 The steps of calculating the elastic coefficient of the road section include: Determine the PM corresponding to the traffic section where the city street view image is located 2.5 The fluctuation ratio in pollution and the PM corresponding to the traffic section where the urban street view image is located 2.5 Recovery ratio after contamination; Calculate and determine the PM 2.5 The recovery ratio after pollution is related to the PM 2.5 The seventh ratio between the fluctuation ratios in pollution is taken as PM 2.5 The road section elasticity coefficient is used to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Elasticity coefficient of road section.

7. The method for predicting urban traffic atmospheric particulate matter concentration according to any one of claims 1 to 6, characterized in that: The vehicle-mounted camera device includes one or more of a driving recorder, a roof camera, a front camera, a rear camera or a side camera; the basic network architecture of the image segmentation model is U 2 net model; the basic network architecture of the atmospheric particulate matter concentration prediction model is a machine learning model, and the machine learning model includes a random forest model.

8. A device for predicting urban traffic atmospheric particulate matter concentration, characterized in that: include: The data acquisition module is configured to respond to the urban traffic atmospheric particulate matter concentration prediction instruction, obtain multiple city street scene images continuously shot by the vehicle-mounted camera device and their corresponding geographical location information, PM concentration of the traffic section where they are located, and 2.5 Concentration, wherein the geographic location information includes one or more of POI information, normalized vegetation index, and building density; a mask data extraction module, configured to use an image segmentation model that has been trained to a convergent state to perform image segmentation on the urban street view image, so as to extract mask data corresponding to the contour of a target area in each target area image in the urban street view image, wherein the target area includes a green plant area, a motor vehicle area, a sky area, a building area, a fence area, a column area, and a wall area; An environmental feature determination module is configured to calculate and determine the number of pixels corresponding to the target area contour in each target area image in the urban street view image according to the mask data, and calculate and determine the urban environmental feature factor according to the number of pixels corresponding to the contour of each target area, wherein the urban environmental feature factor includes a green view rate, a vehicle interference index, a sky visibility index, and a space enclosure index; The elastic coefficient determination module is configured based on the PM 2.5 The concentration calculation determines the PM corresponding to the traffic section where the urban street view image is located. 2.5 Fluctuation rates in pollution and PM 2.5 Recovery ratio after pollution, based on the PM 2.5 Fluctuations in pollution and the PM 2.5 The recovery ratio after pollution is calculated to determine the PM corresponding to the traffic section where the urban street view image is located. 2.5 Section elasticity coefficient; PM 2.5 The concentration prediction module is configured based on the geographical location information, urban environment characteristic factors, fluctuation ratio, recovery ratio, PM 2.5 Section elasticity coefficient and its corresponding PM 2.5 The concentration of the target city is used to construct a sample data set, and the atmospheric particulate matter concentration prediction model trained with the sample data set is used to predict the PM2.5 concentration of the target city corresponding to the traffic section where the street view image is located. 2.5 concentration to complete the prediction of urban traffic atmospheric particulate matter concentration.

9. An electronic device, comprising a central processing unit and a memory, characterized in that: The central processing unit is used to call and run the computer program stored in the memory to execute the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: It stores a computer program implemented according to the method described in any one of claims 1 to 7 in the form of computer-readable instructions, and when the computer program is called and executed by a computer, the steps included in the corresponding method are executed.

Images